PDFs and Text Files

3sparks Chat can send document content to a model so you can summarize, review, transform, or ask questions about it.

Text files

Use the + menu to add text, Markdown, code, JSON, CSV, YAML, and similar files. The app previews file content and estimates token usage before sending.

PDFs

When you add a PDF, the app extracts text and evaluates extraction quality. If text extraction is poor, you can use a vision workflow that converts PDF pages to images for a vision-capable model.

Persistence

File and PDF metadata can remain visible in chat history so you can understand what was sent, even after returning to a conversation later.

Privacy

Document content is sent to the selected provider when you send the message. For private local processing, use Local Models when the content fits your device and model capabilities.

Troubleshooting

• If text extraction looks wrong, try the image/vision path.
• If a file exceeds token limits, send a smaller section.
• If the model ignores the file, ask a specific question about the content.
• If PDF conversion is slow, use fewer pages or a smaller PDF.